This study addresses the relationship between the elaborate structure of the mammalian cone photoreceptor synapse and signaling in postsynaptic bipolar cell retinal pathways that are critical for visual perception. Cone photoreceptors mediate vision during daylight and use glutamate as their transmitter. One of the great insights into early visual processing is that cones contact two functional classes of bipolar cells, On and Off, which have different glutamate receptors that cause them to depolarize (i.e., signal) either at light-on or -off, respectively. Each functional class of bipolar cell can be subdivided into 5-6 anatomical types (cb1a/b, b2, cb3a/b, for Off bipolar cells). Recent work has shown that the different anatomical subtypes of Off bipolar cells 1) express different glutamate receptors (AMPA or kainate), 2) contact the cone terminal at distinct distances from transmitter release sites (in invaginations or at the terminal base), and 3) make different numbers of contacts with a cone terminal. These differences suggest that each type of Off bipolar cell may be specialized to carry a different component of the cone signal to ganglion cells, the retinal output neurons. However, our understanding of how receptor type, contact distance, and contact number lead to differences in Off bipolar cell signaling is far from complete. The first and second aims address the impact of contact number on signaling during conditions of steady light. The basic idea is that vesicle fusion at each of a cone's 20-40 synaptic ribbons is probabilistic, with a mean rate set by membrane voltage. Off bipolar cells make between 2-15 contacts with a cone terminal depending on type. The first aim will use simultaneous cone and Off bipolar cell recording to test key ideas about resampling of the cone signal at the synapse. Experiments will determine the fraction of a cone's ribbon sites sampled by each bipolar cell type, and whether individual contacts at the base of the terminal sample more ribbons than contacts within invaginations. The second aim will test the idea that the more ribbons a bipolar cell samples, the more accurate (or quicker) is its estimate of the cone signal. However, this accuracy comes at the metabolic cost of maintaining more synapses, and thus retinal pathways that don't require high accuracy or temporal precision may likely be fed by Off bipolar cells that make fewer cone contacts. The third aim examines the mechanisms of signaling at light-off, which also differs among the Off bipolar cell types. Cone depolarization at light-off produces a synchronous fusion of vesicles. We hypothesize that cone synapses with cb2 cells saturate during synchronous fusion, providing a mechanism to enhance rapid signaling of change within the visual scene. The ability to rapidly signal change can have utility for survival. We also hypothesize that cb1a cells make specialized contacts with cones that have a threshold non-linearity which may provide noise immunity. Specific experiments will determine the mechanisms of transient signal enhancement at the cone to cb2 and the threshold non-linearity at the cone to cb1a cell synapses.